Article

Robust deep learning object recognition models rely on low frequency information in natural images

2022-02-02

Abstract excerpt

<h4>ABSTRACT</h4> Machine learning models have difficulty generalizing to data outside of the distribution they were trained on. In particular, vision models are usually vulnerable to adversarial attacks or common corruptions, to which the human visual system is robust. Recent studies have found that regularizing machine learning models to favor brain-like representations can improve model robustness, but it is u...

Topics

Open a Topic to create a Post that cites this publication.

Identifiers and source

Literature Corpus work: 1cbc8501-59a8-5a8d-86b5-32c2b0097615
DOI: 10.1101/2022.01.31.478509

Open publication

Related research

Semantic proximity does not establish scientific evidence.

Click a neighbor to travelStep 1 · 12 closest

Robust deep learning object recognition models rely on low frequency information in natural imagesDOI 10.1101/2022.01.31.478509Read

Simulating a Primary Visual Cortex at the Front of CNNs Improves Robustness to Image PerturbationsSemantic proximity Read

Generation and Countermeasures of Adversarial Examples on Vision: A SurveySemantic proximity Read

Beyond AI: Fooling DNN Model in Physical WorldSemantic proximity Read

Adversarial Robustness of Capsule Networks for Medical Image ClassificationSemantic proximity Read

Improving Transferability of Adversarial Examples with Mixed-Representation AttackSemantic proximity Read

Human Sensitivity to Perturbations Constrained by a Model of the Natural Image ManifoldSemantic proximity Read

Chained Dual-GAN: A Generalized Defense against Adversarial AttacksSemantic proximity Read

Color-channel Adversarial Attack with Resolution based CamouflagingSemantic proximity Read

The attentive reconstruction of objects facilitates robust object recognitionSemantic proximity Read

Exposure to naturalistic occlusion promotes generalized, human-like robustness in deep neural networksSemantic proximity Read

Probing Human Categorization of Naturalistic Images by Combining Deep Neural Networks with Cognitive ModelsSemantic proximity Read

Biological convolutions improve DNN robustness to noise and generalisationSemantic proximity Read